Anna Karenina Strikes Again: Pre-Trained LLM Embeddings May Favor High-Performing Learners
This highlights a potential bias in educational AI tools that could disadvantage learners who do not provide correct answers, making it an incremental but important finding for educators and researchers.
The study investigated the use of pre-trained LLM embeddings for clustering student responses in biology, finding that these embeddings poorly captured most expert-defined Knowledge Profiles, except for those with correct answers, revealing a bias in the representation.
Unsupervised clustering of student responses to open-ended questions into behavioral and cognitive profiles using pre-trained LLM embeddings is an emerging technique, but little is known about how well this captures pedagogically meaningful information. We investigate this in the context of student responses to open-ended questions in biology, which were previously analyzed and clustered by experts into theory-driven Knowledge Profiles (KPs). Comparing these KPs to ones discovered by purely data-driven clustering techniques, we report poor discoverability of most KPs, except for the ones including the correct answers. We trace this "discoverability bias" to the representations of KPs in the pre-trained LLM embeddings space.